Method and apparatus for in-band discovery of capabilities of managed devices

FIELD

0001This disclosure relates to management of storage devices and in particular to in-band management of storage devices.

BACKGROUND

0002A storage area network (SAN) is a high-speed network that interconnects different types of storage elements with servers and provides a shared storage pool for servers (hosts) connected to the SAN. The storage elements may include storage arrays, switches, expanders, volume managers, Host Bus Adapters (HBAs) and Redundant Arrays of Independent Disks (RAID).

0003A Redundant Array of Independent Disks (RAID) combines a plurality of physical hard disk drives into a logical drive for purposes of reliability, capacity, or performance. Thus, instead of multiple physical hard disk drives, an operating system sees the logical drive.

0004The Storage Management Initiative-Specification (SMI-S) is a storage standard developed and maintained by the Storage Networking Industry Association (SNIA) and ratified as American National Standards Institute (ANSI) standard ANSI International Committee For Information Technology Standards (INCITS) 388-2004. SMI-S is an object orientated, platform independent standard for managing storage elements in a SAN based upon a Common Information Model (CIM) and Web-Based Enterprise Management (WBEM) standards (using Extensible Markup Language (XML) over Hyper Text Transfer Protocol (HTTP)) defined by the Distributed Management Task Force (DTMF). SMI-S also provides security and automated discovery of storage elements coupled to a SAN. SMI-S allows storage elements to be managed through out-of-band connections.

0005Management of a storage array is typically performed out-of-band over a local area network separate from the SAN. Devices such as the storage array may advertise their presence and manageability capabilities using the Service Location Protocol (SLP) which is an Internet Engineering Task Force (IETF) defined mechanism to advertise the capabilities of any managed device. An SLP template (SLP formatted data) enables a device to publish the availability of any service including manageability, and associated capabilities/attributes to directory agents. Management clients discover the managed devices by querying a directory agent. Discovery involves the retrieval of attributes such as Transmission Control Protocol (TCP)/Internet Protocol (IP) address and port number of the managed device to connect to, which is available in the SLP formatted data registered by the managed device with the directory agent.

BRIEF DESCRIPTION OF THE DRAWINGS

0006Features of embodiments of the claimed subject matter will become apparent as the following detailed description proceeds, and upon reference to the drawings, in which like numerals depict like parts, and in which:

0007<figref idref="DRAWINGS">FIG. 1</figref> is a block diagram of a system that includes a host and a redundant array of independent disks coupled to a storage area network;

0008<figref idref="DRAWINGS">FIG. 2</figref> is a block diagram of a Small Computer Systems Interface (SCSI) Inquiry command;

0009<figref idref="DRAWINGS">FIG. 3</figref> is a block diagram illustrating an embodiment of the format of Inquiry data returned by a target device in response to the SCSI Inquiry command;

0010<figref idref="DRAWINGS">FIG. 4</figref> is a block diagram illustrating in-band management and out-of-band management of an external Redundant Array of Independent Disks (RAID) array in a Small Computer Systems Interface (SCSI) domain;

0011<figref idref="DRAWINGS">FIG. 5</figref> is a block diagram illustrating an embodiment of an SLP template that includes a plurality of attributes;

0012<figref idref="DRAWINGS">FIG. 6</figref> is a flowchart illustrating an embodiment of a method for in-band discovery of device management capabilities according to the principles of the present invention; and

0013<figref idref="DRAWINGS">FIG. 7</figref> is a block diagram of an embedded controller that performs in-band discovery of device management capabilities.

0014Although the following Detailed Description will proceed with reference being made to illustrative embodiments of the claimed subject matter, many alternatives, modifications, and variations thereof will be apparent to those skilled in the art. Accordingly, it is intended that the claimed subject matter be viewed broadly, and be defined only as set forth in the accompanying claims.

DETAILED DESCRIPTION

0015In an embodiment of the present invention, manageability capabilities of a storage device stored as Service location Protocol (SLP) formatted data in the storage device may be retrieved using an in-band mechanism.

0016<figref idref="DRAWINGS">FIG. 1</figref> is a block diagram of a system <b>100</b> that includes a host <b>102</b> and a redundant array of independent disks (RAID) array <b>108</b> coupled to a storage area network (SAN) <b>104</b>. The host <b>102</b> includes a host bus adapter (HBA) <b>110</b>, a Central Processing Unit (CPU) <b>120</b>, a network interface controller (NIC) <b>118</b>, and memory <b>112</b> capable of storing a client <b>114</b> and an operating system (OS) <b>116</b>. The OS <b>116</b> manages a file system that manages files which may be stored in the RAID array <b>108</b>. The client <b>114</b> is a process that issues requests for service through the HBA <b>110</b> to the RAID array <b>108</b>. The HBA <b>110</b> manages communication between the host <b>102</b> and devices such as the RAID array <b>108</b> through the SAN <b>104</b>.

0017The system shown in <figref idref="DRAWINGS">FIG. 1</figref> includes two hosts <b>102</b>, and each host <b>102</b> may access the RAID array <b>108</b> through the SAN <b>104</b>, that is, in-band. Each of the hosts <b>102</b> may also access the RAID array <b>108</b> through a local area network (LAN) <b>106</b>, that is, out-of-band. In-band and out-of band accesses differ in that in-band accesses are directed through the HBA <b>110</b> in the host <b>102</b> using a storage protocol and out-of-band access are directed through the NIC <b>118</b> in the host <b>102</b> over the LAN <b>106</b> using a Transmission Control Protocol (TCP)/Internet Protocol (IP) stack. Out-of-band access is typically used to manage the RAID array <b>108</b>, that is, a separate network (LAN <b>106</b>) is maintained for management access and control. The RAID array may be managed in-band, that is, control and management data share the same network (SAN <b>104</b>) as the data being processed.

0018The RAID array <b>108</b> has front end ports connected to hosts <b>102</b> through the SAN <b>104</b> and backend ports connected to hard disk drives <b>124</b> managed by an array controller <b>122</b>.

0019The RAID array <b>108</b> includes an array controller <b>122</b> and a plurality of hard disk drives (hard drives) <b>124</b>. The array controller <b>122</b> manages the plurality of hard disk drives <b>124</b> which may be aggregated into a single logical storage unit. As is well known to those skilled in the art, there are many standard methods for distributing data across a logical storage unit. These methods are referred to as RAID levels. The array controller <b>122</b> may distribute data across the set of hard disk drives based on one of a plurality of RAID levels.

0020For example, in RAID level 0 data is striped across the physical array of hard disk drives by breaking data into blocks and writing each block to a separate hard disk drive. Input/Output (I/O) performance is improved by spreading the load across many hard disk drives. However, RAID level 0 does not provide redundancy, that is, if one hard disk drive fails all data is lost. Thus, RAID level 0 improves I/O performance but does not provide redundancy.

0021RAID level 5 provides a high level of redundancy by striping both data and parity information across at least three hard disk drives. Data striping is combined with distributed parity to provide a recovery path in case of failure.

0022The SAN <b>104</b> is a high-speed network that is dedicated to providing a high-speed connection between hosts <b>102</b> and storage elements. The hosts <b>102</b> and storage elements communicate over the SAN <b>104</b> using one of a plurality of Small Computer System Interface (SCSI) protocol suites, for example, Internet Small Computer Systems Interface (iSCSI), Serial Small Computer Systems Interface Protocol (SSP) and Fibre Channel Protocol (FCP).

0023A version of the Fibre Channel standard is described in the American National Standards Institute (ANSI) Standard Fibre Channel Framing and Signaling Interface-2 (FC-FS-2) Aug. 9, 2005 Specification. A version of the Fibre Channel Protocol (FCP-3) standard which defines a mapping protocol for applying the Small Computer System Interface (SCSI) command set to Fibre Channel is described in Information technology—Fibre Channel Protocol for SCSI, Third Version (FCP-3) Revision 4, Sep. 13, 2005 American National Standards Institute (ANSI) (hereinafter termed the “FCP standard”). A version of the Serial Attached Small Computer System Interface (SAS) protocol is described in “Information Technology—Serial Attached SCSI—1.1,” ANSI International Committee For Information Technology Standards (INCITS) 417-2006 (International Organization for Standardization(ISO)/International Electrotechnical Commission (IEC) 14776-151)(hereinafter termed the “SAS Standard”).

0024T10 is an ANSI body responsible for developing SCSI Standards that include SCSI Architecture Model (SAM), SCSI Primary Commands (SPC), SCSI Block Commands (SBC), Serial Attached SCSI (SAS), SES (SCSI Enclosure Standard), and SCC (SCSI Controller Commands). SCC Enables RAID Management and SEC Enables Enclosure Management for Storage Arrays. SCSI Addressing methods are described in the ANSI SCSI Architecture Model (SAM) standard.

0025A version of the iSCSI protocol is described in “Internet Small Computer System Interface”, Request for Comments (RFC) 3720 available from the Internet Engineering Task Force (IETF) at www.ietf.org (hereinafter termed the “iSCSI standard”). A version of the SCSI protocol for block-oriented direct-access devices such as magnetic disk drives is described in ANSI INCITS 351-2001 “SCSI Primary Commands-2 (SPC-2) and ANSI INCITS 405-2005 “SCSI Block Commands-2 (SBC-2)”. iSCSI encapsulates the SCSI protocol in an Internet Protocol (IP) packet that is transmitted using the IP protocol over the network.

0026The serial attached storage protocols provide a connection-orientated class of service between devices. Typically, in a serial attached storage protocol, a connection is established between an initiator (originator) and a target (responder). The initiator may be a storage protocol controller such as a Host Bus Adapter (HBA) and the target may be a storage device, for example, a hard disk drive, Digital Video Disk (DVD) drive, compact disk (CD) drive, Redundant Array of Independent Disks (RAID), or tape drive.

0027After the connection is established between the initiator and the target, commands, data and status information encapsulated in frames are exchanged between the initiator and the target. Frames may be received in the same order that they are transmitted. A frame is a package of information transmitted as a single unit. The format of the frame and encapsulated information is defined by the protocol suite.

0028A SCSI storage element also referred to as a SCSI logical unit includes a logical unit number (LUN). The logical unit number which is stored in non-volatile memory in the SCSI logical unit is used to persistently identify the SCSI logical unit. Typically, 64-bits (8 bytes) are allocated for storing the logical unit number. Although 64-bits are allocated, all of the bits may not be used for all storage protocols.

0029The logical unit number may be accessed from the SCSI logical unit during the discovery process through the use of a SCSI Report LUN command which is a SCSI primary command and is defined in the ANSI T10 SCSI standards. Then the SCSI Inquiry command is used to obtain basic information from a target device and allows a host to discover the basic information associated with the target device. The SCSI Inquiry command may be used to determine the configuration of a logical unit with the logical unit responding with information that includes type, version of the standard used, vendor's identification, and model number. This information may be stored on the medium or in non-volatile memory in the target device.

0030<figref idref="DRAWINGS">FIG. 2</figref> is a block diagram illustrating the format of a SCSI Inquiry command. The SCSI Inquiry command is sent from a SCSI initiator to a SCSI target to request that the SCSI target return basic information associated with the SCSI target, for example, information about SCSI commands supported by the SCSI target.

0031The SCSI Inquiry command <b>200</b> shown in <figref idref="DRAWINGS">FIG. 2</figref> is a six byte command that includes an operation code <b>202</b>, a page code <b>210</b>, an allocation length <b>214</b> and control <b>216</b>. The operation code <b>202</b> identifies the SCSI command to be a SCSI Inquiry command. The state of the Enable Vital Product Data (EVPD) field <b>208</b> indicates whether vital product data specified by the page or operation code field <b>210</b> is requested. If the EVPD field <b>208</b> is zero, standard Inquiry data is returned in response to the SCSI Inquiry command.

0032<figref idref="DRAWINGS">FIG. 3</figref> is a block diagram illustrating the format of an embodiment of Inquiry data <b>300</b> returned by a target device in response to the SCSI Inquiry command. The Inquiry data <b>300</b> includes at least 36 bytes (bytes [<b>35</b>:<b>0</b>]) of standard Inquiry data <b>302</b>. The first 36 bytes in the standard Inquiry data <b>302</b> includes the total number of bytes returned by the Inquiry command. The standard Inquiry data <b>302</b> also includes the peripheral device type, vendor identification, product identification, and product revision level of the target device. The Inquiry data <b>300</b> may include vendor specific data <b>304</b> stored in bytes <b>55</b>:<b>36</b> and bytes N:<b>96</b>.

0033The vendor specific data field <b>308</b> is used to store management capabilities of the target device. The management capabilities will be described later in conjunction with <figref idref="DRAWINGS">FIG. 5</figref>.

0034<figref idref="DRAWINGS">FIG. 4</figref> is a block diagram illustrating in-band management and out-of-band management of an external RAID array <b>108</b> (<figref idref="DRAWINGS">FIG. 1</figref>) in a Small Computer Systems Interface (SCSI) domain. The management of the system <b>100</b> shown in <figref idref="DRAWINGS">FIG. 1</figref> includes a Storage Management Initiative-Specification (SMI-S) management domain and a storage domain as shown in <figref idref="DRAWINGS">FIG. 4</figref>.

0035SMI-S defines an object oriented message based interface for controlling elements in a storage area network. SMI-S capabilities are defined by a hierarchy of object classes with properties and methods. The object oriented interface allows a client to discover, monitor, and control storage devices irrespective of the implementation of the storage devices.

0036In the SMI-S domain, identifiers of Common Information Model (CIM) instances are used to identify storage elements. In the storage domain, storage identifiers are used to identify storage elements. In one embodiment, the storage identifiers are SCSI logical unit number (LUN) identifiers defined by the SCSI standards. The SMI-S object model is based on the CIM. Web Based Enterprise Management (WBEM) also referred to as CIM/Extensible Markup Language (XML) over Hyper Text Transfer Protocol (HTTP) implements the messaging based interface. WBEM Clients are management clients that communicate with a Common Information Model Object Manager (CIMOM) using the CIM/XML/HTTP protocol suite.

0037In an embodiment, the SMI-S management domain includes a management client <b>400</b>, a Common Information Model Object Manager (CIMOM) <b>402</b> and a management provider <b>404</b> in the host <b>102</b> of <figref idref="DRAWINGS">FIG. 1</figref>. The storage domain includes a device driver <b>406</b>, a RAID Storage Controller <b>408</b> and RAID targets <b>410</b> in the RAID array <b>108</b> of <figref idref="DRAWINGS">FIG. 1</figref>. The management provider <b>404</b> performs a discovery process which involves the enumeration of all SLP targets and their management capabilities communicates with the storage target <b>410</b> through the HBA <b>110</b> (<figref idref="DRAWINGS">FIG. 1</figref>), and the device driver <b>406</b>.

0038In another embodiment, the CIMOM <b>402</b> and management provider <b>404</b> may be in the RAID array <b>108</b> of <figref idref="DRAWINGS">FIG. 1</figref>.

0039The Service Location Protocol (SLP) is an IETF standards-track protocol that simplifies the discovery and use of network resources such as printers, web servers, fax machines, and storage devices such as RAID and tape drives. SLP is a packet-oriented protocol with packets transmitted using User Datagram Protocol (UDP) or Transmission Control Protocol (TCP).

0040The Service Location Protocol (SLP) allows devices to announce services on a local network allowing computers and other devices to find services in the local area network without the need for prior configuration. In the embodiment shown, a RAID array <b>410</b> includes an SLP service agent <b>418</b> that allows the RAID array <b>410</b> to announce services to an SLP directory agent <b>416</b>.

0041The SLP service agent <b>418</b> includes at least one service. The service includes a Uniform Resource Locator (URL) that is used to locate the service. The service also includes a plurality of name/value pairs, called attributes. A service template includes the definition of a service, URL and allowed attributes. The service template is defined by IETF Request For Comments (RFC) 2609 and SLP is defined in Service Location Protocol, Version 2, both of which are available at www.ietf.org.

0042Instead, of providing the name of a network host (a human readable text string that is an alias for a network address) supporting a service, a description of the service (type of service and a set of attributes which describe the service) may be provided, for example, by an administrator (user) of the system through the management client <b>410</b>. Based on the description, the Service Location Protocol resolves the network address of the service (through the management client).

0043The client <b>114</b> (<figref idref="DRAWINGS">FIG. 1</figref>) uses the SMI-S standard to manage storage elements such as the RAID array <b>108</b> (<figref idref="DRAWINGS">FIG. 1</figref>) in the SAN <b>104</b> (<figref idref="DRAWINGS">FIG. 1</figref>) using one or more objects defined by the SMI-S standard for the storage element. The SMI-S model is divided into several profiles. A profile describes a class of SAN entities, for example, a switch or a disk array. For example, SMI-S defines an array profile instance to describe an external RAID array.

0044The storage domain includes a device driver (HBA driver) <b>406</b> which is a device specific control program that enables operation with a particular device. In one embodiment, the device driver <b>406</b> enables operation with the RAID storage controller <b>408</b>. The RAID storage controller <b>408</b> manages an array of hard disk drives in the RAID <b>410</b> using one of a plurality of serial storage protocols, for example, Serial Attached Small Computer System Interface (SAS) or Fibre Channel (FC).

0045A management client <b>400</b> is a process that runs in any one of the hosts <b>102</b> shown in <figref idref="DRAWINGS">FIG. 1</figref>. In an embodiment, the management client <b>400</b> may enable an administrator to configure and control the managed elements in the storage domain. The management client <b>400</b> issues management requests to perform operations on the managed elements. The management requests may identify the storage elements using their assigned SMI-S/CIM instance identifiers and identify the operations to be performed using SMI-S defined methods. For example, the management client <b>400</b> may be client <b>114</b> in host <b>102</b> shown in <figref idref="DRAWINGS">FIG. 1</figref>.

0046In an embodiment, the management client <b>400</b> may be a Web-Based Enterprise Management (WBEM) client that communicates with a Common Information Model Object Manager (CIMOM) <b>402</b> using the CIM/XML/HTTP protocol suite through a messaging interface. Through the messaging interface, the management client <b>400</b> issues management requests to the CIMOM <b>402</b>.

0047The managed elements include a RAID array <b>410</b> which may be represented as a Common Information Model (CIM) object. The management client <b>400</b> communicates with the CIMOM <b>402</b> which manages CIM objects. The CIMOM <b>402</b> transfers information to a managed element through the management provider <b>404</b>. The management provider <b>404</b> includes software drivers that interface managed elements with CIMOM <b>402</b>, and enable their control using the management client <b>400</b>.

0048In an embodiment, the CIMOM <b>402</b> is in the host <b>102</b> shown in <figref idref="DRAWINGS">FIG. 1</figref>. In response to a message received from the management client <b>400</b>, the CIMOM <b>402</b> generates an SMI-S method for a CIM instance address associated with the managed element which is forwarded to the management provider <b>404</b>.

0049Upon receiving a management request from the CIMOM <b>402</b>, the management provider <b>404</b> performs the discovery process by issuing commands through the device driver <b>406</b> to the RAID array <b>410</b>. One such command may be to obtain configuration data for the RAID array <b>410</b> which may be provided by the SLP service agent <b>418</b> in the RAID array <b>410</b>.

0050The SLP service agent <b>418</b> includes SLP formatted data (SLP template) per storage device. For example, in the RAID array <b>410</b> shown in <figref idref="DRAWINGS">FIG. 4</figref>, there is one SLP template stored for each storage device in the RAID array. The SLP Formatted data is available both in-band through the SCSI Inquiry command and out-of-band through the SLP Protocol. An example of an SLP formatted data is shown in <figref idref="DRAWINGS">FIG. 5</figref>. In the embodiment shown, the SLP formatted data for a SCSI storage array (RAID array <b>410</b>) includes the address of the SCSI LUN that responds to in-band SMI-S commands, and the address of the SCSI LUN that generates asynchronous SCSI events.

0051In the embodiment shown, each storage element in the RAID (storage array) <b>410</b> has a respective logical unit number. The logical unit number may be used to uniquely identify each storage element in the RAID array <b>410</b>.

0052The management provider <b>404</b> also stores information for each storage element used by the device driver <b>406</b> to address the managed elements in an array of storage identifiers. This information includes a SCSI logical unit number (LUN) identifier associated with each managed element that uniquely identifies the storage element.

0053The management provider <b>404</b> uses the information retrieved from each logical unit number in the returned Inquiry data to manage the storage element in the SCSI domain.

0054The in-band transportation of the SCSI Inquiry data that includes the SLP service template is delivered to the management provider <b>404</b> through the device driver <b>406</b> and the RAID storage controller <b>408</b> over path <b>414</b> which may use a serial storage protocol such as Fibre channel or SAS to the SCSI target device in the RAID array <b>410</b>.

0055The out-of-band mechanism for managing storage devices such as the external storage array is through the WBEM protocol and Ethernet physical links over path <b>412</b>. The target device advertises its presence and manageability capabilities to the SLP directory agent <b>416</b>. Management clients <b>400</b> discover the managed devices by querying the SLP directory agent <b>416</b>. Discovery involves the retrieval of attributes such as TCP/IP address and port number of the managed device and this information is available in the SLP formatted data registered by the managed device with the SLP directory agent <b>416</b>.

0056In the in-band mechanism, the SLP formatted data returned in the SCSI Inquiry response by the SLP service agent <b>418</b> may be stored by the management provider <b>404</b> in the SLP directory agent <b>416</b>. Thus, the manageability capabilities retrieved from the target device using the in-band mechanism is available for managing the target device by the management provider <b>404</b>.

0057The SLP template is a formal description of the SLP attributes and service scheme associated with SLP. The various attributes of the SLP Template are used to describe the in-band management capabilities of the SLP target device. Attributes are used to advertise the management capabilities of the target device. In one embodiment, the SLP template is returned in the SCSI Inquiry response in XML format. In anther embodiment the SLP template is in American Standard Code for Information Interchange (ASCII) format.

0058In another embodiment, the SLP formatted data (SLP template) may include both in-band and out-of-band attributes allowing all management attributes to be retrieved both in-band and out-of-band. Out-of-band attributes that may be included in the SLP template include the TCP/IP port that the target device listens on and the version of the web management application supported by the target device. In-band attributes that may be included in the SLP formatted data (SLP template) include the logical unit number through which the service is available. Thus, a device capable of both in-band and out-of-band management can advertise its capability in a synchronized and standardized manner.

0059In one embodiment, the attributes included in the SLP formatted data returned in the SCSI Inquiry command response data may include attributes for advertising a WBEM server. These attributes may include attributes that are defined in the “WBEM SLP template”, version 1.0 available at www.dmtf.org/standards/wbem/wbem.1.0.en

0060<figref idref="DRAWINGS">FIG. 5</figref> is a block diagram illustrating an embodiment of an SLP template <b>550</b> that includes a plurality of attributes. The WBEM SLP template includes a template type <b>500</b> that includes the service type name, which may be ‘inband-wbem’ and a template version <b>502</b> that provides the version number of the service type specification. The SLP template <b>550</b> includes attributes that are used to describe both in-band and out-of-band management capabilities of a target device. The SLP template may be registered with the directory agent <b>416</b> (<figref idref="DRAWINGS">FIG. 4</figref>) by the target device RAID <b>410</b> or the management provider <b>404</b> (<figref idref="DRAWINGS">FIG. 4</figref>).

0061The command Logical Unit Number (LUN) field <b>504</b> specifies the 8 byte address of the LUN through which the ‘inband-wbem’ service is available. Although the target device is discovered through the SCSI Inquiry command, it is possible that the ‘inband-wbem’ service may be available at a LUN other than LUN 0. The event LUN field <b>506</b> specifies the 8-byte address of the LUN through which the ‘inband-wbem’ event service is available. The SCSI LUN stored in the command LUN field <b>504</b> is the LUN to which in-band management commands are sent, for example, in-band SMI-S commands. The event LUN field <b>506</b> stores the SCSI LUN that responds to SCSI commands such as, a SCSI Enclosure Services Command. In one embodiment the command LUN field <b>503</b> and the event LUN field <b>506</b> store different LUN numbers.

0062The WBEM server attributes may also include name and description of a CIM service for a human interface and an identifier for the WBEM server. The SLP template <b>550</b> also includes a communication mechanism (protocol) field <b>508</b> and other communication mechanism field <b>510</b> for storing the communication mechanisms used by a CIM object manager (CIMOM <b>402</b> (FIG. <b>4</b>)), for example, XML or ASCII text format and a protocol Version field <b>512</b> for storing a version of the protocol. The functional profiles supported field <b>514</b> and functional profiles description fields <b>516</b> describe and list the CIM operation profiles and operations supported by the CIM object manager. The namespace field <b>520</b> is for storing namespaces supported on the target device. The class information field <b>522</b> is provided for storing class information (classes), that is, CIM schema version supported for the name spaces listed in the namespace attribute field <b>520</b>. The registered profile supported field <b>524</b> indicates the profiles supported by the target device. The classes and profiles may be out-of-band attributes, that is, the same SLP template may describe both in-band and out-of-band attributes. For example, the SLP template may also include other out-of-band attributes such as TCP/IP address and port number.

0063The WBEM server registers its service and the attributes that accompany its service. A client application may display the service information in a user interface, for example, using the human readable information provided by the attributes.

0064Thus, the in-band mechanism for determining management capabilities of a managed element may be the similar to the out-of-band mechanism. That is, in-band discovery of device management capabilities is similar to the out-of-band discovery by retrieving the SLP formatted data in the response data returned from the SCSI Inquiry command. The need for a proprietary vendor unique implementation for in-band discovery of manageability capabilities is avoided through the use of standard SCSI supported and WBEM supported mechanisms. Furthermore, a single management provider may query the management capabilities of all devices that include SLP formatted data in the SCSI Inquiry command response data.

0065The management provider <b>404</b> (<figref idref="DRAWINGS">FIG. 4</figref>) may register devices that may only be managed in-band with the directory agent. For example, referring to <figref idref="DRAWINGS">FIG. 1</figref> an embodiment of the target device <b>108</b> shown in <figref idref="DRAWINGS">FIG. 1</figref> without a Network Interface Card (NIC) <b>118</b> could only be managed in-band through the storage area network <b>104</b> because there would be no access through the local area network <b>106</b>. A service registration includes a service: URL and optionally a set of attributes.

0066The management provider <b>404</b> (<figref idref="DRAWINGS">FIG. 4</figref>) publishes the SLP formatted data as it is received from the target, that is, no translation of the SLP formatted data is required. Also, management clients <b>400</b> (<figref idref="DRAWINGS">FIG. 4</figref>) may discover devices that are managed both in-band and out-of-band.

0067<figref idref="DRAWINGS">FIG. 6</figref> is a flowchart illustrating an embodiment of a method for in-band discovery of device management capabilities according to the principles of the present invention. <figref idref="DRAWINGS">FIG. 6</figref> will be described in conjunction with <figref idref="DRAWINGS">FIGS. 2-5</figref>.

0068At block <b>600</b>, the management provider <b>404</b> requests capabilities of a discovered target device. A request to return capabilities of the target is issued to the SCSI domain which results in issuing a SCSI Inquiry command <b>200</b> (<figref idref="DRAWINGS">FIG. 2</figref>) to the target device. In one embodiment, the standard SCSI Inquiry command <b>200</b> described in conjunction with <figref idref="DRAWINGS">FIG. 2</figref> may be used to discover the target device with the EVPD field <b>208</b> set to 1, the page or operation code field <b>210</b> set to C0 (hexadecimal) and the allocation length field <b>214</b> set to 96 bytes. If the target response data is more than 96 bytes, another SCSI Inquiry command may be issued with the appropriate allocation length. Processing continues with block <b>602</b>.

0069At block <b>602</b>, the management provider <b>402</b> (<figref idref="DRAWINGS">FIG. 4</figref>) waits for SCSI Inquiry response data <b>300</b> (<figref idref="DRAWINGS">FIG. 3</figref>) to be returned. The SCSI Inquiry response data <b>300</b> is returned to the SCSI domain from the target device and forwarded to the management provider <b>404</b> if the SCSI Inquiry command completes successfully. In one embodiment all of the SCSI Inquiry response data <b>300</b> may be returned to the management provider <b>404</b>. The detection of the successful completion of the SCSI Inquiry command is handled by the SCSI domain and is beyond the scope of the present invention. Upon receiving SCSI Inquiry response data after a successful completion has been detected by the SCSI domain, processing continues with block <b>604</b>.

0070At block <b>604</b>, the management provider <b>404</b> receives the SLP formatted data <b>550</b> (<figref idref="DRAWINGS">FIG. 5</figref>) that has been extracted from the vendor specific data field <b>308</b> of the SCSI inquiry response data <b>300</b>. In another embodiment the management provider <b>404</b> may extract the SLP formatted data from the returned SCSI Inquiry response data if all of the SCSI Inquiry response data is forwarded from the SCSI domain to the management provider <b>404</b>. Processing continues with block <b>606</b>.

0071At block <b>606</b>, the management provider <b>404</b> stores the extracted SLP formatted data <b>550</b> in the SLP directory agent <b>416</b> for use by the management domain. Processing continues with block <b>608</b>.

0072At block <b>608</b>, the management provider <b>404</b> determines if there is another device for which management capabilities are to be retrieved. If so, processing continues with block <b>600</b>. If not, processing is complete.

0073An embodiment of the invention has been described for a host bus adapter (HBA) in a computer system. An embodiment of the invention may also be provided in an embedded controller for a drive array as shown in <figref idref="DRAWINGS">FIG. 7</figref>. The drive array includes a plurality of storage devices which may be accessed through SAS or SATA drive ports via the SAS expander <b>710</b>. A host interface <b>702</b> provides communication to a host system and a network interface controller <b>704</b> provides support for out-of-band management of the embedded controller. The host interface <b>702</b> may include the management provider <b>404</b>, SLP directory agent <b>416</b> and SLP service agent <b>418</b> discussed in conjunction with <figref idref="DRAWINGS">FIG. 4</figref>.

0074In one embodiment the network interface controller <b>704</b> is a 1 Giga bit (Gb) Ethernet controller. The memory <b>706</b> is coupled to a SAS/RAID controller <b>708</b> for storing data used for implementing the RAID functionality.

0075It will be apparent to those of ordinary skill in the art that methods involved in embodiments of the present invention may be embodied in a computer program product that includes a computer usable medium. For example, such a computer usable medium may consist of a read only memory device, such as a Compact Disk Read Only Memory (CD ROM) disk or conventional ROM devices, or a computer diskette, having a computer readable program code stored thereon.

0076While embodiments of the invention have been particularly shown and described with references to embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of embodiments of the invention encompassed by the appended claims.